(Adapted from the application) The broad, long-term objective of the proposed research is to provide structural information important for the development of peptide-based antigens that illicit immune responses protective against HIV infections. The information will be obtained by structural characterization of HIV peptides bound to class II MHC proteins and by structure-based combinatorial design of tight-binding synthetic peptide antigens. 1) To determine the mode of binding of "non-conforming" HIV-derived antigenic peptides. The t-cell response to HIV infection includes several immunodominant peptide epitopes that appear to violate the MHC peptide binding motifs determined by biochemical and structural studies. Determination of the three-dimensional structure of the human class II MHC protein HLA-DR1 in complex with "non-conforming" peptide epitopes from HIV will help characterize any novel features of their mode of binding. 2) To determine the structure of the complex of HLA-DR1 with the HIV env- derived antigenic peptide gp120[304-318]. Determination of the structure of HLA-DR1 complexes of serine and arginine variants of a gp120 that is a strong class II mhc-restricted T-cell antigen and is contained within the hypervariable V3-loop that is the primary target of neutralizing antibodies will help to characterize the immune response to gp120 and the viral mechanisms that evade the host response. 3) To design tight-binding peptide analogues that function as synthetic T-cell antigens with improved properties. Design efforts will focus on retaining conserved MHC-peptide main-chain interactions while varying the peptide backbone and side chains to maximize interactions between the peptide and pockets within the overall MHC peptide binding cleft. Three dimensional structures of peptide complexes will be used to guide the design of highly biased combinatorial libraries that will be screened for high affinity binding and long dissociating half-life.